Re: Dropping Ethernet Connections [1 Attachment]

"petermcs-/[email protected] [rabbit-semi]" <[email protected]>
Newsgroups gmane.comp.hardware.rabbit-semiconductor
Message-ID <[email protected]>
I don't know if it is the same in DC9 but...
 

 In DC 10, the ARP library tries to do opportunistic caching of addresses when it receives broadcasts on the basis that it might come in useful. The only problem with this is that the ARP cache is usually very small and it ends up flushing useful addresses from the cache to replace them with addresses that may not be any use at all. I've modified my libraries to not do this opportunistic caching as I found it led to a lot of ARP traffic for addresses that I'd already used but had been flushed.
 

 I suspect a combination of this and some sort of bug in the ARP handling is causing your issue. One possible solution is to increase the ARP cache size.
 

 You might look at the attached file line 764 onwards for what I did with DC10 UDP library and see if it is applicable to the DC 9 libs.
 

 Regards,
 Peter
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.